Five patients with sucrase-isomaltase deficiency, and one patient with primary glucose-galactose malabsorption had no increases in breath hydrogen excretion after oral sucrose or glucose. Anaerobic incubation with sugars of stool suspensions from 5 patients with primary sugar malabsorption produced by trace of hydrogen (17 microliter) in only one, while those from 13 or 14 controls produced a mean hydrogen volume of 640 microliter under similar conditions. Altered bacterial metabolism is a probable explanation. Breath hydrogen excretion did increase appreciably in 2 of these patients after oral lactulose showing that hydrogen excretion may vary according to the substrate. Therefore, observation of breath hydrogen excretion after lactulose is not recommended as a means of predicting false-negative breath tests with other sugars. The hydrogen breath test is not a reliable mean of diagnosing primary sugar malabsorption in children.